Toxic substances associated with particles from a variety of sources accumulate in fresh and saltwater sediments where they provide long- term exposures to humans and ecosystems. Given the nationwide efforts at industrial source control, contaminant loadings to receiving waters have decreased. Contaminants present within sediments caused historical exposures and in some cases, current exposures to humans and ecosystems. It is the aim of this research to develop and test a methodology for reconstructing historical exposures to trace metals in estuarine systems ain order to anticipate human disease endpoints that may require 20 to 40 years to materialize. Field studies are planned at two sites that are typical of what is encountered throughout coastal sites in the United States. One site has very simple geometry of a constructed marina and has well defined activities from a military base that have contribute contaminants since 1940. A two-year study at this site can develop a complete understanding of tidal flows, the seasonal cycle of freshwater input, and the connection between water column processes and contaminant burial. The other site is more complex spatially and chemically and will require the remaining three years of effort to test this methodology for historical exposure assessment. The second site is on the margins of San Francisco Bay where tidal flows over mud flats were altered by construction activities over the last 60 years. There are multiple sources of trace metal contamination from industrial and mineral processing activities. The combined efforts will be devoted to a historical analysis of activities, measuring the depth distribution of contaminants and tracers over the site, and undertaking hydrodynamic measurements in the water column. These data provide the input to a model that will determine the contaminant levels in the near surface sediments for prior years and predict associated exposure to humans through fish consumption. This effort that historical exposure assessment and dose reconstruction will optimally combine monitoring data of existing conditions, chemical characterization of the site, and mathematical modeling for predicting historical conditions. The modeling effort will explicitly account for uncertainties in model predictions. Through this effort, historical exposures of humans to contaminants in estuarine systems are assessed and possible health needs for previous exposed humans can be anticipated.